Search bioRxiv⌕ Search

Biology subjects

Cirrincione, G.

Publications and source records attributed to Cirrincione, G..

2 recordsLinked to original sources

Self-Architecting Protein Transformers: An Empirical Study

Motivation. Protein language models (pLMs) such as ESM-2 and ProtBERT rely on pretraining corpora of tens to hundreds of millions of sequences and on encoder architectures whose depth, width and number of attention heads are chosen by the practitioner and never revisited during training. The entry cost of state-of-the-art pLMs is therefore out of reach for laboratories without industrial-scale infrastructure, and the fixed architecture provides no in-training diagnostic of whether the chosen capacity matches the structural complexity of the data. This work asks whether a self-architecting transformer, which grows its own width and depth from quantitative signals derived from the attention matrices, can extract competitive protein representations from a single reference proteome. Results. A three-level self-architecting framework, INCRT-geo, is applied to masked-language pretraining on the human Ensembl proteome (approximately twenty thousand sequences). On Pfam-50 family classification, the principal model attains a linear-probe accuracy that exceeds two pretrained baselines, ESM-2 small and ProtBERT, despite a corpus several orders of magnitude smaller. Three single-variable ablations isolate the contributions of one-residue tokenisation, depth growth and an asymmetry-loss regulariser; the regulariser is shown to be necessary for the depth-growth trigger to fire. Scaling pretraining to eight vertebrate proteomes does not improve Pfam accuracy under the available compute budget; the negative result is reported transparently. Architectural diagnostics indicate that the heads allocated by the framework are functionally diverse rather than redundant.

bioinformatics↗

Exploring the therapeutic potential of a novel series of imidazothiadiazoles targeting focal adhesion kinase (FAK) for pancreatic cancer treatment: Synthesis, mechanistic Insights and promising antitumor and safety profile

Focal Adhesion Kinase (FAK) is a non-receptor protein tyrosine kinase that plays a crucial role in various oncogenic processes related to cell adhesion, migration, proliferation, and survival. The strategic targeting of FAK represents a burgeoning approach to address resistant tumors, such as pancreatic ductal adenocarcinoma (PDAC). Herein, we report a new series of twenty imidazo[2,1-b][1,3,4]thiadiazole derivatives assayed for their antiproliferative activity against the National Cancer Institute (NCI-60) panel and a wide panel of PDAC models. Lead compound 10l exhibited effective antiproliferative activity against immortalized (SUIT-2, CAPAN-1, PANC-1, PATU-T, BxPC-3), primary (PDAC-3) and gemcitabine-resistant clone (PANC-1-GR) PDAC cells, eliciting IC50 values in the low micromolar range (1.04-3.44 {micro}M), associated with a significant reduction in cell-migration and spheroid shrinkage in vitro. High-throughput kinase arrays revealed a significant inhibition of the FAK signalling network, associated to induction of cell cycle arrest in G2/M phase, suppression of tumor cell invasion and apoptosis induction. The low selectivity index/toxicity prompted studies using PDAC mouse xenografts, demonstrating significant inhibition of tumor growth and safety. In conclusion, compound 10l displayed antitumor activity and safety in both in vitro and in vivo models, emerging as a highly promising lead for the development of FAK inhibitors in PDAC.

cancer biology↗